Placental peptides metabolism and maternal factors as predictors of risk of gestational diabetes in pregnant women. A case-control study

Background Gestational diabetes is a risk factor for perinatal complications; include shoulder dystocia, birth injuries such as bone fractures and nerve palsies. It is associated with later development of type 2 diabetes, the risk of macrosomia and other long-term health effects of infants born to diabetic mothers. The study assesses placental peptides and maternal factors as potential predictors of gestational diabetes among pregnant women. Material and methods A total of 200 pregnant women were recruited for the study, 150 pregnant women without pre gestational diabetes including 50 women with low risk factors of diabetes as controls and 50 other pregnant women with pregestational diabetes as control. Fasting blood glucose and the lipid profile were determined by enzymatic methods using Envoy® 500 reagents (Vital Diagnostics, USA). Glycated haemoglobin was assessed using the Cation Exchange resin method. Leptin and the Human Placenta Lactogen were assayed using the Sandwich-ELISA technique. Beta chorionic gonadotrophin, insulin, progesterone and estradiol were determined using chemilumiscence imunoassay technique on MAGLUMI 600 analyzer. Anthropometry, including BMI and blood pressure were also measured. Results Fasting plasma glucose (FBG), insulin, insulin resistance, glycated haemoglobin and Human Placenta Lactogen(HPL)were significantly (p<0.0001) increased in the pregestational diabetic women whereas progesterone and estradiol were significantly decreased. In the second trimester however, there was no significant difference (p>0.05) in estradiol, insulin, insulin resistance and HPL between the pregnant women who developed gestational diabetes and those who did not. Leptin, progesterone and FBG were significantly increased in those who developed GDM. The risk of developing gestational diabetes increased with overweight (OR = 1.76, P = 0.370) and family history of diabetes (OR = 2.18, P = 0.282). Conclusion Leptin, progesterone, estradiol estimated in this study were increased in the gestational diabetes mellitus women and fairly predicted gestational diabetes in the non-diabetics pregnant women. Obesity, aging and family history of diabetes were strongly predictive of gestational diabetes.


Introduction
In a normal pregnancy, sensitivity to insulin reduces with increasing gestation age [1].It is a physiological programming designed to shift metabolic fuel to the fetus for development [2]. In this state, insulin sensitivity in pregnant non-diabetics is reportedly reduced [3].The decrease in insulin sensitivity with progression in pregnancy is due partly to the inhibitory effect of placenta peptides (C-reactive protein, leptin and human placental lactogen etc) on insulin secretion [4,5].The effects of these placenta peptides and hormones cause the enlargement of the islets of Langerhans cells and or the hyperplasia of the pancreatic β-cells [6] to increase the secretion of more insulin, resulting in compensated hyperinsulinaemia [7]. Obesity and chronic insulin resistance during pregnancy are most common factors that exacerbate β-cell dysfunction [8]. Gestational diabetes mellitus (GDM) occurs when the pancreatic β-cells are unable to produce sufficient insulin to offset the persisting insulin resistance in pregnant women [3].
Similar frequencies of HLA-DR2, DR3 and DR4 antigens in GDM as in healthy pregnant women have been reported and this has been found to be associated with low prevalence of autoimmune destruction of β-cells and therefore rules out the possibility that GDM is a disease of autoimmune origin [9].
Gestational diabetes mellitus is reported in 2 to 9 percent of all pregnancies, [10] it is a risk factor for adverse perinatal complications that may later result in the development of type 2 diabetes. [11]. The long-term health effects on infants born to mothers with gestational diabetes include impaired glucose tolerance, [12] obesity [13] and impaired intellectual ability [14].
Pregnant women with high risk factors for GDM or showing clinical signs of diabetes are usually subjected to an oral glucose tolerance test (OGTT) at first trimester and then again between 24 and 28 weeks if the initial tests were negative. At weeks 24-28 of gestation, it is recommended that all women undergo an OGTT. GDM is defined as a fasting plasma glucose level !5.5 mmol/l and/or a 2-h plasma glucose level !8.5 mmol/l after a fasting 75-g OGTT [15]. Treatment of gestational diabetes reduces perinatal morbidity and may also improve the woman's health-related quality of life [16].
Screening for GDM is limited, the standard screening procedure for diabetes and pre-diabetes, such as fasting blood glucose and glycated haemoglobin, are not recommended for screening for GDM. At first trimester FBG alone does not distinguish between GDM, impaired glucose tolerance or preexisting diabetes, also HbA1c predicts glucose metabolic state over a period of 3 months and therefore does not distinguish between preexisting diabetes and GDM. However, the recommended method is an oral glucose tolerance test (OGTT) [17] and this is expensive, invasive, and requires hospital visit and multiple blood draws.
The development of abnormal glucose metabolism after gestational diabetes can be predicted readily by means of available clinical and anthropometric measurements. [18,19] Similarly, if there were a proper assessment of the various risks as well as markers that may predict the onset of gestational diabetes among pregnant women, it would in turn allow practitioners to better manage patients and prevent many pregnancy associated complications. The current study therefore was aimed at assessing the roles of placental peptides and maternal factors as potential predictors of gestational diabetes among pregnant women in the Tema Metropolis of Ghana.

Methods
A cross-sectional case-control study was carried out at the Tema General Hospital, Tema Polyclinic and Provita Specialist Hospital all in Ghana, from March 2014 to March 2015. A total of 200 pregnant women aged between 17-45 years and mean age of 28.43±4.95 years were recruited for the study, comprising of 150 pregnant women without pregestational diabetes (subjects) including 50 women in the group with low risk factors of diabetes as controls. Fifty (50) other pregnant women with pregestational diabetes were recruited as controls.

Study participant selection
To determine gestational diabetes in pregnant women, a standard OGTT was performed between the 24 th and 28 th week of gestation after an overnight fast (8-12 hours), by administering 75 g anhydrous glucose in 300 ml water. Women were advised to follow a normal diet 48 hours before the oral glucose-tolerance test (OGTT). However, pregnant women belonging to high risk diabetes population were screened during the first trimester in order to detect a previously undiagnosed diabetes mellitus. Gestational age was also determined on the basis of the woman's last normal menstrual period and if it coincided within 1week of the date determined by ultrasound done between 16 th and 20 th weeks of gestation, otherwise the ultrasound estimates was used. A pretested guided questionnaire was administered to collect demographic data of family history of diabetes, parity, miscarriages, and complications of pregnancy. Anthropometric parameters including BMI and blood pressures were measured. Pregnant women who reported at the antenatal clinic were included; unless they opted out of the study, non pregnant women seeking fertility advice were excluded from the study.

Informed consent of subjects
Clients who reported at the antenatal clinic and agreed to participate in the study gave written informed consent after detailed explanation of the study procedures had been given to them before the recruitment into the study. Guardians of participants below 18 years of age gave consent on their behalf. glucose) for analysis. Blood samples were allowed to clot. After clotting, samples were centrifuged at 3000 g for 4 min and the sera separated into appropriate tubes and frozen at -20˚C until analysis. Metabolites analyzed for included total cholesterol, triglyceride (TG), low-density lipoprotein (LDL), high-density lipoprotein (HDL), fasting blood glucose serum insulin, glycosylated haemoglobin, leptin, estradiol, progestrone, Human placenta lactogen and beta HCG. The homeostasis model assessment index-insulin resistance (HOMA-IR), based on fasting insulin and glucose measured in a single blood sample, was used to calculate insulin resistance [20].

Biochemical analysis
The frozen serum samples were thawed to room temperature before being analyzed.

Lipid profile
Triglycerides, total cholesterol and HDL-cholesterol were assayed using enzymatic methods with Envoy 1 500 reagents (Vital Diagnostics, USA) according to the manufacture's specification on BT 5000 1 Random Access Chemistry Analyzer (Biotecnica, Italy).

Glycated haemoglobin
The Cation Exchange Resin method for the selective estimation of glycated haemoglobin (HbA1c) method in human whole blood was used, on Microtech 3000 semi auto analyzer.

Leptin
Maternal serum leptin concentration was measured quantitatively by Sandwich-ELISA technique using Leptin (Human LEP) kit (Elabscience Biotechnology co ltd WuHan P.R.C). The lowest limit of detection of the assay was 0.094ng/mL. The intra assay and inter assay CV ranged from 2.5% to 9.5% and from 4.7% to 8.3%, respectively.

Human placenta lactogen (HPL)
Maternal serum Human placenta lactogen concentration was measured quantitatively by Sandwich-ELISA technique on Human Reader HS, a semi auto analyzer.

Progestrone (P4) and estradiol (E2)
Progestrone and Estradiol, were determined using chemilumiscence imunoassay technique on MAGLUMI 600 analyzer. The method uses ABEI label anti-E2/anti-prog antibody and a purified E2/ Prog antigen to coat magnetic microbeads. The ABEI labeled microbeads were incubated at 37˚C forming antigen antibody complex which was precipitated in a magnetic field and the supernatant descanted. The addition of starter I and 2 initiates chemilumiscence reaction, and the light signal produced measured.

Beta chorionic gonadotrophin (β-HCG) and Insulin assay
Beta chorionic gonadotrophin and insulin were also determined using MAGLUMI 600 analyzer. The sandwich chemilumiscence imunoassay technique was used, adhering to the manufacturers' protocol.

Statistical analysis
Results were expressed as mean ± S.D and statistical analysis performed using SPSS version 20.0 (SPSS Inc.) and GraphPad prism 5 for Windows. Normal distribution and homogeneity of the variances were tested using Kolmogorov-Smirnov and Levène tests, respectively. Student t-test was used to compare the significance of the difference in the mean values of any two groups and chi square analysis was used to compare frequency between the two groups. Linear regression analysis was used to study the correlation between the parameters. Correlations between parameters were analyzed using the Pearson R test for variables with normal distribution and ROC to determine the sensitivity and specificity of FBG, HbA1c, Insulin and β-HCG as markers of GDM. P<0.05 was considered statistically significant. Table 1 show 50 women with pre-gestational diabetes (controls) and 150 pregnant women without diabetes (cases) during first trimester, 50 of which have very low risk of developing diabetes. Fasting plasma glucose, insulin, insulin resistance, glycated haemoglobin, and Human Placenta Lactogen were significantly increased (p<0.0001) in the pregestational diabetic controls, whereas, progesterone and estradiol were significantly increased in the nondiabetic pregnant women. BMI was also significant (p<0.05) increased in pregnant women with pregestational diabetes (Table 1). After screening for gestational diabetes among the pregnant women without diabetes (n = 150), between 24-28 week gestation,12 of them developed GDM giving a prevalence of 8% GDM. Fasting plasma glucose, glycated haemoglobin, leptin, and progesterone were significantly (p<0.0001,p = 0.029, p<0.0001,p = 0.043 respectively) increased in the pregnant women who developed GDM. Estradiol and HPL were non significantly changed in the GDM women, whilst plasma haemoglobin was higher in the women without GDM (Table 2).

Results
When the pregnant women were stratified according to age, pregnant women less than 20yrs of age did not present with GDM, whilst, 33.3% of them in the age range of 20-29yrs had GDM. The highest prevalence of 58.3% GDM occurred at 30-39yrs, whilst only 8.3% GDM was found in the 40-49yrs range (Fig 1) Pregnant women who developed GDM had significantly (p<0.01) raisedfasting blood glucose, (p<0.014), BMI (P = 0.022),leptin, progesterone, estradiol (p<0.0001) than the pregnant women with low risk of diabetes (Table 3)but significantly low(p<0.0001) human placenta lactogen levels.  Table 4 shows the Accuracy of diagnosis of first trimester FBG-1, FBG-2, HbA1c, Insulin, β-HCG, P4, E2 and HPL.
The diagnostic value of 3.75 mmol/L for first trimester fasting blood glucose (FBG-1) had a sensitivity of 33.3% and Specificity of 72.0% as marker for GDMand5.6 mmol/L diagnostic value with significant (p<0.0001) 100% sensitivity and 95.7% specificity for second trimester fasting blood glucose (FBG-2). First trimester HbA1c with sensitivity of 33.3% and specificity of 87.0% had a threshold value of 5.6%. β-HCG showed 100% sensitivity and specificity of 7.65% at 4947. E2 and HPL showed similar sensitivity of 50.0% but difference in specificity of 63.0% and 71.7% at 5256 and 65.61 respectively.
The risk of developing GDM increased with age (OR = 1.02, P = 0.757). Overweight (OR = 1.76, P = 0.370) and family history of diabetes (OR = 2.18, P = 0.282) ( Table 5). Pregnant women in the first trimester, with a history of abortion, miscarriage and pre-mature birth as complications had higher risk of developing GDM. Also the risk of developing GDM increased with increasing plasma progesterone.

Discussion
Pregnancy has been identified as a diabetogenic condition characterized by the development of insulin resistance and hyperinsulinemia [7]. It has been proposed that pregnancy-induced insulin resistance reveals the onset of β-cell defects associated with insulin resistance and hyperglycemia, hyperlipidaemia [21]. which underlie GDM. GDM is diabetes restricted to pregnant women in whom the development of glucose intolerance was first diagnosed during pregnancy, distinguishing it from preexisting glucose intolerance and diabetes induced by pregnancy [21,22]. This study was aimed at assessing placental peptides and maternal factors as potential predictors of the development of gestational diabetes among pregnant women.
The pre-gestational diabetic pregnant women had significantly higher plasma glucose, insulin and insulin resistance than the non diabetic pregnant women ( Table 1).The plasma glucose and insulin resistance in the non diabetic pregnant women were significantly low probably due to the early stage of pregnancy. However, it has been shown that insulin resistance is associated with normal pregnancy and may not develop into GDM [23]. The mechanisms leading to insulin resistance in normal pregnancy are explained by several theories including; the estrogen and progesterone hormone counter effect on insulin action by their inhibitory effect on the skeletal muscle β-subunit of the insulin receptor and insulin receptor substrate-1 [23] and thereby making available enough glucose for the growing foetus. [24]. These changes in insulin signaling reduces insulin-mediated glucose uptake in the muscle, the major tissues for whole-body glucose disposal [25]. However, insulin resistance is terminated soon after delivery and normal glucose metabolism returns within 1 year postpartum [26]. This phenomenon of insulin resistance in normal pregnancy was observed in this study. There was no significant difference, in plasma insulin, insulin resistance, β-HCG (P = 0.436)) and the lipid profile between pregnant women diagnosed with GDM and pregnant women with no diabetes at the early weeks of gestation (Table 2). This seems to suggest that, these peptides and the lipids may not play significant roles in the aetiology of GDM.
However, as the gestation progresses the effect of placenta peptides; Progestrone, estradiol, leptin and glycated haemoglobinon insulin predisposes the women to glucose intolerance and eventually to overt GDM [27,28,29]. Progestrone, and estradiol were significantly (p<0.0001) increased in the cases compared to the controls (Table 1). Progesterone, estradiol and leptin were also significantly higher in women with GDM than women with low risk of developing GDM (Table 3). This phenomenon may account for the role of these placenta peptides in the aetiology of GDM. Women with GDM have a combination of acquired and chronic insulin resistance and are more insulin resistant compared to the normal women during late pregnancy. The phosphorylation of insulin receptor tyrosine is implicated in the transmission of insulin signal to enable glucose uptake 30 . In women with GDM, there is a significant decrease in insulin receptor tyrosine phosphorylation in the skeletal muscle, and a decrease in insulinstimulated glucose uptake [30]. Therefore, pregnancy can unmask slight defects that induce GDM. It has been proposed that these peptides interfere with insulin phoshorylation during pregnancy and impair glucose uptake [29]. The mean plasma leptin in the non diabetics pregnantwomen was nonsignificantly raised (Table 1), probably due to the smaller number of the women (12/150) who developed GDM had high leptin levels. This assertion is verified in Table 3, where leptin levels in GDM women is significantly higher than in women with low risk of GDM. Leptin concentration in amniotic fluid at 15-17 weeks of gestation is higher in women who subsequently developed GDM than in women with normal glucose tolerance during pregnancy [31]. Leptin enhances peripheral insulin sensitivity and improves pancreatic β-cell function [32].These normal functions of leptin are disrupted in leptin resistance state of pregnancy inducing insulin resistance and impaired glucose tolerance in pregnancy [33,34]. The placenta produces leptin during pregnancy resulting in further hyperleptinaemia. This is exacerbated by hyperleptinaemia due to fat accumulation, and subsequent insulin resistance, giving rise to GDM in women who are unable to compensate for the physiological dysregulation.
Anthropometric changes have been implicated in the aetiology of GDM. BMI was significantly (p = 0.022) higher in GDM women compared to pregnant women with low risk of GDM (Table 3). The risk of developing gestational diabetes increased with overweight (OR = 1.76, P = 0.370) ( Table 5). Similar findings reported by Chu et al., shows that the risk of developing GDM is about 2 and 4 times higher among overweight and obese women respectively, as compared to normal-weight pregnant women [35].
Aging is a major confounding factor associated with the development of GDM. Even though the odd ratio of developing GDM due to ageing was low (OR = 1.02, P = 0.757) ( Table 5), women within the age range of 30-39 had the highest probability of developing GDM (58%), whilst those at 20yrs and below had the lowest probability of 0.0%. (Fig 1). As women advance in age, there is a tendency to increased BMI, adiposity, multiparity and high blood pressure, and these are all confounding factors of GDM [36].
Apart from age, and BMI, other confounding factors that predisposes pregnant women to the development of GDM, include, family history of diabetes, ethnicity, adiposity and parity [37]. In a North American population, higher rates of GDM were reported among blacks, Hispanics, and Asians compared to non-Hispanic whites [38,39]. A family history of diabetes increased the odds of developing GDM (OR = 2.18, P = 0.282). Pregnant women with a history of abortion, miscarriage and pre-mature birth as complications had higher risk of developing GDM ( Table 5).The risk of developing GDM increased with increasing plasma insulin (OR 1.04 P = 0.256) and progesterone (OR 1.04 P = 0.05). Also the placental peptides were significantly higher in women who developed GDM than those who had low risk for the development of GDM (Table 3). Progestrone also had a strong sensitivity and specificity in diagnosing GDM, whilst β-HCG, estradiol and HPL were poor predictors of GDM.
Glycated haemoglobin, fasting blood glucose and fasting plasma insulin, are the key markers of diabetes. None of the placenta peptides and hormones had a significant specificity and selectivity as a marker of GDM in the first trimester. However, plasma glucose of diagnostic value of 5.6mmol/l had significant specificity and selectivity as marker of GDM in the second trimester (Table 4).
Body mass index was significantly and positively correlated with glycated haemoglobin (.702 Ã ). Fasting blood glucose had an inverse and strong significant correlation (-.801 ÃÃ ) with insulin levels among GDMs (Table 6). Though, HbA1c, insulin and β-HCG are useful metabolites for diagnoses and monitoring of diabetes mellitus, yet had fair diagnostic tests values for GDM, since the AUC (95% CI) were >0.05. Glycosylated haemoglobin is not usually used clinically to diagnose GDM, it is however useful in diagnosing pre-existing diabetes in early pregnancy [40].
The area under the ROC for FBG-1 (at first trimester) was 0.49 suggesting that fasting plasma glucose level in the first trimester is a poor diagnostic test for gestational diabetes mellitus. However, AUC for FBG-2 was 0.99 implying that FBG-2 is a good diagnostic test for diagnosing GDM between 24wks-28 wks, whereas, glycated haemoglobininsulin and β-HCG were poor diagnostic test markers for GDM as compared to FBS-2 (Fig 2).

Conclusion
The crude prevalence of GDM in the metropolis was 8.0%.The plasma level of the placenta peptides, Leptin, Progesterone and Estradiol estimated were increased in the women who developed gestational diabetes mellitus. These peptides were fair in predicting GDM. However, progesterone level in the first trimester was outstanding in predicting the onset of GDM. Underweight, overweight, age and history of diabetes were strong predictors of GDM. HbA1c, insulin and β-HCG are fair tests for predicting GDM.

Limitations
The effect of some confounding factors, ethnicity, adiposity on gestational diabetes were not independently determined.